proteomic analysis of leaf proteins during dehydration of the resurrection plant xerophyta viscosa

Proteomic analysis of leaf proteins during dehydration ofthe resurrection plant Xerophyta viscosa

ROBERT A. INGLE1, ULRIKE G. SCHMIDT2, JILL M. FARRANT1, JENNIFER A. THOMSON1 &SAGADEVAN G. MUNDREE1

1Department of Molecular and Cell Biology, University of Cape Town, Private Bag, Rondebosch 7701, South Africa and2Department of Molecular Plant Physiology, University of Zurich, Zollikerstrasse 107, CH-8008, Zurich, Switzerland

ABSTRACT

The desiccation-tolerant phenotype of angiosperm resur-rection plants is thought to rely on the induction of pro-tective mechanisms that maintain cellular integrityduring water loss. Two-dimensional (2D) sodium dodecylsulphate–polyacrylamide gel electrophoresis (SDS–PAGE)analysis of the Xerophyta viscosa Baker proteome wascarried out during dehydration to identify proteins that mayplay a role in such mechanisms. Quantitative analysisrevealed a greater number of changes in protein expressionlevels at 35% than at 65% relative water content (RWC)compared to fully hydrated plants, and 17 dehydration-responsive proteins were identified by tandem mass spec-trometry (MS). Proteins showing increased abundanceduring drying included an RNA-binding protein, chloro-plast FtsH protease, glycolytic enzymes and antioxidants. Anumber of photosynthetic proteins declined sharply inabundance in X. viscosa at RWC below 65%, including fourcomponents of photosystem II (PSII), and Western blotanalysis confirmed that two of these (psbP and Lhcb2) werenot detectable at 30% RWC. These data confirm thatpoikilochlorophylly in X. viscosa involves the breakdown ofphotosynthetic proteins during dismantling of the thylakoidmembranes. In contrast, levels of these photosyntheticproteins were largely maintained during dehydration inthe homoiochlorophyllous species Craterostigma plan-tagineum Hochst, which does not dismantle thylakoidmembranes on drying.

Key-words: desiccation tolerance; photosynthesis; poikilo-chlorophylly; proteome.

INTRODUCTION

Desiccation tolerance describes the ability of an organismto survive drying to equilibrium with the relative humidityof air, reviving when water again becomes available (Illinget al. 2005; Alpert 2006). The majority of higher plantsproduce desiccation-tolerant seeds and pollen, but it is arare phenomenon in vegetative tissues (Vicré, Farrant &Driouich 2004a; Illing et al. 2005). However, a small group

of taxonomically diverse plants known as resurrectionplants are able to tolerate severe (> 95%) water loss fromtheir vegetative tissues for prolonged periods, and recoverfull physiological activity upon rehydration (Gaff 1971;Ingram & Bartels 1996; Vicré et al. 2004a). Xerophytaviscosa (Velloziaceae) is one such species, and is able tosurvive drying to 5% relative water content (RWC) andresumes full physiological activity within 80 h of rewatering(Sherwin & Farrant 1998). The origin of desiccation toler-ance in vegetative tissues is polyphyletic (Oliver, Tuba &Mishler 2000) and it has been suggested that it arose viathe activation of seed-specific protection mechanisms invegetative tissue (Oliver et al. 2000; Illing et al. 2005). Theobservation that homologs of two seed-specific genes inArabidopsis thaliana, LEA6 and a 1-Cys peroxiredoxin areexpressed in vegetative tissues of the resurrection plantsXerophyta humilis and Xerophyta viscosa, respectively, atRWC < 65% (Mowla et al. 2002; Illing et al. 2005) lendssome support to this hypothesis.

Desiccation tolerance in plants is thought to rely on twostrategies: the protection of cellular integrity and the repairof dehydration- or rehydration-induced damage (Cooper &Farrant 2002). Lower-order plants such as the moss Tortularuralis which desiccates rapidly (in as little as 1 h) rely onconstitutive damage repair mechanisms during rehydration,and are classified as full desiccation-tolerant plants as tol-erance is unaffected by the rate of drying (Oliver & Bewley1997). In contrast, higher-order plants rely more on theinduction of mechanisms to protect cellular integrity duringwater loss, and are classified as modified desiccation-tolerant plants, as a certain amount of time is requiredfor the induction of tolerance (Oliver & Bewley 1997). Totolerate desiccation, resurrection plants must be able tocombat three major subcellular stresses; mechanical stress,e.g. shrinkage of the plasma membrane away from the cellwall and subsequence cytorhesis, oxidative stress causedby the production of reactive oxygen species (ROS) anddamage to macromolecules such as DNA and proteins(Farrant 2000; Alpert 2006). Mechanisms employed byplants to prevent mechanical damage include changes incell wall composition to allow folding (Vicré et al. 2004b;Moore et al. 2006), cytoplasmic packaging with vacuoles inwhich water is replaced with compatible solutes (Farrant2000;VanderWilligen et al. 2004) and changes in membrane

Correspondence: R. A. Ingle. Fax: +27 21 689 7573; e-mail:[email protected]

Plant, Cell and Environment (2007) 30, 435–446 doi: 10.1111/j.1365-3040.2006.01631.x

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd 435

lipid composition to increase fluidity (Moore et al. 2005).The inability to utilize light energy in photosynthesis duringwater stress can lead to the elevated production of ROS(Smirnoff 1993). Resurrection plants carry out a controlledshutdown of photosynthesis early on during drying (atRWC > 50%) to avoid the generation of ROS, and alsoemploy a range of antioxidant systems to limit damage(Sherwin & Farrant 1998; Farrant 2000). Poikilochloro-phyllous species such as X. viscosa dismantle thylakoidmembranes and degrade chlorophyll, while homoiochloro-phyllous species retain the majority of their chlorophyll andrely on mechanisms such as anthocyanin production andleaf folding to prevent chlorophyll–light interactions(Sherwin & Farrant 1998). Strategies to protect the integrityof macromolecules include the synthesis of compatiblesolutes such as sucrose and trehalose (Müller, Boller &Wiemken 1995; Ingram et al. 1997; Whittaker et al. 2001)and the synthesis of late embryogenesis abundant (LEAs)and chaperone proteins (Bernacchia & Furini 2004).

To identify genes that may underlie such desiccation tol-erance mechanisms, a number of studies have examinedchanges that occur during dehydration and/or rehydrationin the transcriptome of several resurrection plants. Theseinclude X. humilis (Collett et al. 2003, 2004), Craterostigmaplantagineum (Bartels et al. 1990; Bernacchia, Salamini &Bartels 1996; (Bockel, Salamini & Bartels 1998) andSporobolus stapfianus (Blomstedt et al. 1998; Neale et al.2000) and the moss T. ruralis (Wood, Duff & Oliver 1999;Oliver et al. 2004). However, there has been little corre-sponding study of the proteome of any resurrection plant.De novo protein synthesis during rehydration has previ-ously been studied in both T. ruralis and C. plantagineumusing two-dimensional (2D) sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) (Oliver& Bewley 1984; Bernacchia et al. 1996), but the identity ofthe newly synthesized proteins was not determined. In addi-tion, several dehydration-induced proteins were recentlyshown to be transiently phosphorylated in C. plantagineumduring a dehydration using 2D SDS–PAGE coupled with aphosphoprotein specific stain (Röhrig et al. 2006). Massspectrometry (MS) analysis in conjunction with the increas-ing amount of protein sequence data [over 250 000sequences in the National Center for Biotechnology Infor-mation (NCBI) Viridiplanta database] now offers theopportunity to identify homologous proteins in non-modelplant species such as X. viscosa.A major advantage of pro-teomics over transcriptomics is that it focuses on theactively translated portion of the genome. The importanceof post-transcriptional regulation has been demonstratedby several studies revealing a weak or moderate correlationbetween mRNA and protein levels, except for very abun-dant proteins in yeast (Gygi et al. 1999; Ideker et al. 2001).Afurther consideration in the case of resurrection plants isthat mRNAs appear to be stored during drying and onlytranslated during rehydration (Dace et al. 1998;Collett et al.2003), thus there may be significant differences betweenmRNA and protein levels during dehydration. A study ofthe desert legume Retama raetam provides support for this

hypothesis. This species shuts down photosynthesis in itsupper stems during the dry season (Mittler et al. 2001).However, while protein levels of psbA (the D1 polypep-tide) and large and small subunits of ribulose 1·5-bisphosphate carboxylase/oxygenase (Rubisco) are muchreduced in the upper stems compared to the lower stems,there is no corresponding difference in mRNA levels.Protein levels recovered within 6–24 h after watering sug-gesting that this species may also store mRNAs during dryperiods (Mittler et al. 2001). The aim of the present workwas thus to investigate changes in the proteome that occurin X. viscosa during dehydration to identify proteins thatmay play a role in desiccation tolerance mechanisms in thisspecies.

MATERIALS AND METHODS

Plant material and culture

Xerophyta viscosa Baker plants were collected from Cathe-dral Peak, Kwa-Zulu Natal, and C. plantagineum Hochstplants from the Pilanesberg Nature Reserve, NorthwestProvince, South Africa. The plants were potted and grownunder glasshouse conditions as described by Sherwin &Farrant (1996). Prior to this study, the plants were trans-ferred to a controlled environment room with a photosyn-thetic flux of 350 mmol m-2 s-1 under a 16 h light/8 h darkcycle at 25 °C and allowed to acclimate for 2 weeks beforedehydration stress was imposed.

Dehydration treatment and determinationof RWC

Three independent biological replicate groups of X. viscosaplants, each consisting of three individual plants that weredried down by withholding water. Tissue samples were har-vested from these plants prior to dehydration (100%RWC)and at two points during dehydration (65 and 35% RWC).All samples were collected at noon to avoid apparent dif-ferences in protein abundance caused by circadian or light–dark regulation. Fresh and dry (after oven drying at 70 °Cfor 48 h) biomass was determined gravimetrically for eachleaf sample harvested (10 per plant), and absolute watercontent (AWC) calculated using the following formula:(fresh biomass - dry biomass)/dry biomass. RWC was cal-culated using the following formula: (AWC ¥ 100)/AWC atfull turgor (determined by bagging plants overnight afterwatering).

Isolation of total protein

Total protein for one-dimensional (1D) and 2D SDS–PAGE was isolated from 1 g of leaf tissue as previouslydescribed (Ingle, Smith & Sweetlove 2005).

Two dimensional gel electrophoresis

Total soluble protein (800 mg) was separated by isoelectricfocusing on 18-cm-long immobilized non-linear pH 3.0–10.0

436 R. A. Ingle et al.

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Plant, Cell and Environment, 30, 435–446

gradient strips (GE Healthcare, Bucks, UK) in the firstdimension and SDS–PAGE in the second, and stained withcolloidal Coomassie blue as previously described (Millaret al. 2001). Three replicate gels were run per RWC, eachwith a protein sample isolated from an independent biologi-cal replicate pool of three plants.

Analysis of 2D SDS–PAGE gels

Gel images were acquired by flatbed scanning.Protein spots(averaging 428 � 52 per gel) were detected in these images,quantified and matched across gels using PDQuest software(Bio-Rad, Hercules, CA, USA). Default settings were usedduring automated spot detection and matching, prior tomanual inspection and any necessary editing of spot match-ing. There was no significant difference in the number ofspots detected per gel at the three RWCs. Spot quantitieswere normalized to total valid spot volume on the gel, anddehydration-responsive proteins identified by paired t-testanalysis of spot quantities on the control gels compared tothe corresponding spots at 65 or 35% RWC. Only spotspresent on all three biological replicate gels at all threeRWCs were included in this analysis. Biological variationbetween the plants (which were collected from differentlocalities within the Drakensberg) is likely to be greaterthan any technical variation in the system, and so proteinswere considered to be differentially expressed if the pairedt-test analysis was significant at the P < 0.05 level, i.e. thesame pattern of change was observed for a given protein inall three biological repeats. In addition, a qualitative analy-sis was carried out using PDQuest to detect spots presentonly during dehydration in all three biological replicates at65/35%RWC (‘de novo’ proteins). Proteins of interest wereexcised from the gel using a cut 200 mL pipette tip, andstored in 96-well microtitre plates at -80 °C prior to MSidentification.

MS identification of proteins

Protein spots were digested with trypsin and mixed with asolution of a-cyano-4-hydroxycinnamic acid (5 mg mL-1 ina solution containing 0.03% trifluoroacetic acid and 65%acetonitrile). Approximately 0.7 mL of a 1:3 mixture ofdigest solution and matrix-assisted laser desorption/ionization (MALDI) matrix solution were deposited on a192-well MALDI target plate.MS analyses were performedon a 4700 Proteomics Analyser MALDI time-of-flight(TOF)/TOF system (Applied Biosystems, Framingham,MA,USA).The instrument was equipped with a Nd : YAGlaser operating at 200 Hz.All mass spectra were recorded inpositive reflector mode and were generated by accumulat-ing data from 5000 laser pulses. Up to 5 spectral peaks perspot that met the threshold criteria were included in theacquisition list for the MS/MS spectra. Peptide fragmenta-tion was performed at collision energy of 1 kV, and a colli-sion gas pressure of approximately 1.5 ¥ 10-6 Torr. DuringMS/MS data acquisition, a minimum of 3000 laser pulsesand a maximum of 6000 laser pulses were allowed for each

spectrum. Global Proteomics Server (GPS) Explorer Soft-ware (Applied Biosystems) was used for submitting MSdata for database searching with the Mascot search engine(www.matrix-science.com). All searches were performedagainst the Viridiplanta NCBI database with the followingsettings: maximum number of missed cleavages, 1; peptidetolerance, � 50 ppm; MS/MS tolerance, � 0.2 Da. Car-boxyamidomethylation of cysteine was set as fixed modifi-cation and oxidation of methionine were selected asvariable modification.The criteria for considering a proteinto be successfully identified were: (1) at least two peptideswere matched to the sequence of the homolog in the NCBIdatabase and (2) the overall protein MOWSE score wassignificant at P < 0.05. The molecular weight/isoelectricpoint (MW/pI) predicted for the homologs in the NCBIdatabase was in good agreement with that estimated fromthe SDS–PAGE gels.

Western analysis

Total protein (30 mg) was separated on 12% SDS–PAGEgels and transferred to nitrocellulose membrane overnight.Equal loading of protein was confirmed by Ponceau stain-ing of the membrane. Membranes were blocked for 2 hor overnight at 4 °C in 1 ¥ Tris-buffered saline Tween-20 (TBST) [except psbP where 1 ¥ phosphate-bufferedsaline Tween-20 (PBST) was used] containing 10% w/vnon-fat milk powder. Primary antibodies were diluted in1 ¥ TBST [2-Cys peroxiredoxin (2-Cys-Prx) 1:2000, FtsH1:1000, HCF136 1:1000, Lhcb2 1:2000 and psbS 1:2000] or1 ¥ PBST (psbP 1:300) containing 10% w/v non-fat milkpowder. Blots were incubated with primary antibody for1.5 h, followed by 3 ¥ 5 min washes in 1 ¥ TBST or1 ¥ PBST (psbP) and incubation with secondary anti-body (IgG HRP, 1:5000 dilution) for 1 h. Bands weredetected using chemiluminescence, and sized by compari-son to a prestained protein ladder.

Measurement of ascorbate content

The total ascorbate content was determined from 0.15 g oftissue according to the method of Wang, Jiao & Faust(1991), except that 0.35% (w/v) iodoacetamide rather thanN-ethylmaleimide was used to alkylate excess dithiothreitol(DTT) after reduction of dehydroascorbate to ascorbate.

RESULTS AND DISCUSSION

Leaf samples were taken from X. viscosa plants at fullturgor (100% RWC) and during drying at 65 and 35%RWC, and total soluble protein extracted and analysed by2D SDS–PAGE. These two RWCs represent two distinctphases of the dehydration process where induction of puta-tive ‘early’ and ‘late’ protection mechanisms is initiated(Mundree & Farrant 2000; Illing et al. 2005).At 65% RWC,Xerophyta species are still green and photosyntheticallyactive although at a reduced rate (Farrant 2000), while at35% RWC, photosynthesis has ceased and chlorophyll has

Dehydration responses in the X. viscosa proteome 437


been degraded. Of the approximately 430 protein spotsdetected on the gels, quantitative analysis using PDQuestidentified 20 that were significantly increased in abundance(paired t-test, P < 0.05) during drying, while 13 showed asignificant decrease compared to levels at 100% RWC. Inaddition, Boolean analysis identified 21 protein spotspresent in all three biological replicates at 65 and/or 35%RWC (referred to here as ‘de novo’ proteins) that could notbe detected in soluble protein extracts from fully hydratedplants (Fig. 1). These may represent proteins that areonly produced, or are only present at detectable levelsduring dehydration. However, the possibility that post-translational modification, for example, phosphorylationhas altered theMW/pI of these proteins during dehydrationcannot be ruled out. In all three classes (increased,decreased and de novo proteins), there were a greaternumber of changes at 35% RWC, particularly of proteinsshowing decreased expression. The 54 proteins could bedivided into three groups based on their expression pat-terns. Eight proteins were early-dehydration responsive(significant change only at 65% RWC), suggesting a pos-sible role during the initial stages of drying, while 24 werelate-dehydration responsive (significant change only at35% RWC), perhaps indicating a role in the latter stages ofdrying. The remaining 22 proteins were full-dehydrationresponsive, with altered levels of expression at both 65 and35% RWC (Fig. 1).

These 54 protein spots were cut from the gels for MSidentification. To determine whether peptide mass finger-printing (PMF) was a viable option for protein identifica-tion, eight of these spots, plus spots corresponding to thelarge subunit of Rubisco and the b-subunit of the F1

ATPase (based on size and relative position on the gels),were subjected to MALDI-TOF analysis. These two highlyconserved proteins were successfully identified by PMF;however, a significant score for protein ID was onlyobtained for one of the eight experimental samples aftersearching of the NCBI databases with the Mascot searchengine. This protein corresponded to the 33 kDa oxygen-evolving complex of photosystem II (PSII) (psbO) andshowed reduced expression in X. viscosa at 35% RWC. Toachieve a higher identification rate, trypsin digest frag-ments were subjected to MALDI-TOF/TOF analysis. Thisresulted in a significant score for protein ID being obtainedfor a further 16 of the remaining 53 samples (an ID rate of30%).The putative MS/MS identifications of these proteinsare shown in Table 1, amino acid sequences of the matchedpeptides in Table 2 and representative gels at 100, 65 and35%RWC in Fig. 2.To validate the protein expression data,antibodies were obtained to four of the proteins identifiedby MS/MS: psbP, HCF136 (down-regulated), 2-Cys-Prx andFtsH (up-regulated), andWestern blotting was carried out.The expression patterns determined by immunodetectionwere in good agreement with those determined by 2DSDS–PAGE (Fig. 3).

Changes in expression of photosyntheticproteins during drying in X. viscosa

The abundance of five chloroplast proteins involved in pho-tosynthesis was significantly decreased at 35% RWC: psbOand psbP, two components of the luminal oxygen evolvingcomplex (OEC) of PSII, the PSII stability factor HCF136,the a-subunit of the F-ATPase and the Calvin cycle enzymetransketalose (Table 1). Of these, only HCF136 was alsosignificantly lower at 65% RWC, which correlates with thereported decline of the chlorophyll fluorescence parameterratio of variable/maximum fluorescence (Fv/Fm) indicating areduced rate of electron transport through PSII (Farrant2000; Mundree & Farrant 2000). HCF136 is a thylakoidluminal protein required for PSII stability and assembly;Arabidopsis hcf136 mutants lack PSII activity and do notaccumulate PSII subunits (Meurer et al. 1998). Reducedlevels of HCF136 protein in X. viscosa during drying may beone component of the shutdown of photosynthesis if thisserved to reduce the rate of stable PSII formation. It hasrecently been reported that the OEC may be the primarysite of photoinhibition of PSII, with oxidative damage tothe D1 polypeptide occurring secondarily once the OECis unable to supply electrons to reduce oxidized P680+

(Hakala et al. 2005; Ohnishi et al. 2005). Thus, the decreasein psbO and psbP protein levels in X. viscosa during dryingcould represent either down-regulation or degradationcaused by damage, or both. The observation that psbP andpsbO mRNA levels also decline in the related species Xero-phyta humilis at RWCs below 50% suggests that at leastsome of the observed decrease in protein levels is likelycaused by down-regulation of gene expression (Collettet al. 2003, 2004). Finally, a twofold decrease in levelsof glutamate:glyoxylate aminotransferase I (GGT1) was

De novo

Increased Decreased

21

20 13

2 6132 613

‘Early’ response – 65% RWC only (8)

‘Late’ response – 35% RWC only (24)

2 922 924 974 97

‘Full’ response – 65 and 35% RWC (22)

Dehydration response

Figure 1. Venn diagram illustrating the expression patterns ofdehydration-responsive proteins from Xerophyta viscosa.Proteins were classified into three groups based on theirexpression: early-dehydration responsive [change in expression at65% relative water content (RWC) only], late-dehydrationresponsive (35% only) or full-dehydration responsive (65 and35% RWC). ‘De novo’ protein indicates that the protein was notdetected in soluble protein extracts from fully hydrated plants.



Tab

le1.

Massspectrom

etry

(MS)

iden

tific

ationof

proteins

show

ingasign

ificant

chan

gein

abun

dancein

Xer

ophy

tavi

scos

adu

ring

dehy

dration

Putativeiden

tific

ation

Species

Accession

numbe

rMea

suredMW

/pI

Predicted

MW/p

IRespo

nse

Dec

reas

edab

unda

nce

65%

35%

1Tran

sketalose

Sola

num

tube

rosu

mCAA90

427

78/5.4

80/5.9

nsd

0.41

�0.13

2F-A

TPase(a

subu

nit)

Ran

uncu

lus

mac

rant

hus

AAZ03

784

59/5.2

55/5.3

nsd

0.40

�0.13

3Glu:glyox

ylate

aminotransferase

IA

rabi

dops

isth

alia

naAAN62

332

52/6.5

54/6.5

nsd

0.47

�0.06

4PSIIstab

ility

factor

HCF136

Ory

zasa

tiva

BAD62

115

40/5.2

45/9.0

0.44

�0.1

0.58

�0.15

5Ascorba

tepe

roxida

seA

.tha

liana

CAA66

925

32/5.0

28/5.9

nsd

0.11

�0.01

6PsbO

A.t

halia

naCAA36

675

30/5.6

35/5.7

nsd

0.34

�0.06

7PsbP

Xer

ophy

tahu

mili

sAAN77

240

26/7.0

26/6.9

nsd

0.14

�0.04

Incr

ease

dab

unda

nce

8Chlorop

last

FtsH

protea

seA

.tha

liana

CAA68

141

72/5.2

76/5.8

3.14

�0.77

3.30

�1.50

9GDP-m

anno

se-3

′,5′-e

pimerase

O.s

ativ

aQ2R

1V8

47/6.2

43/5.8

2.18

�0.30

2.67

�0.52

10Alcoh

olde

hydrog

enase

Citr

us¥

para

disi

AAY86

033

46/6.4

42/5.8

nsd

2.22

�0.16

11Protein

phosph

atasetype

2CA

.tha

liana

CAB79

642

33/6.0

30/6.4

4.20

�0.36

3.92

�0.07

12VDAC1.1

Lot

usco

rnic

ulat

usAAQ87

019

28/6.5

30/8.6

nsd

1.81

�0.20

132-Cys

peroxiredo

xin

O.s

ativ

aCAJ016

9326

/5.0

28/5.7

nsd

1.72

�0.52

De

novo

prot

eins

14dn

aK-typ

emolecular

chap

eron

eO

.sat

iva

NP_0

0104

8274

75/5.3

73/5.5

Present

Present

15RNA-binding

protein

Dau

cus

caro

taAAK30

205

76/6.3

72/6.9

Absen

tPresent

16Pho

spho

pyruvate

hydratase

Zea

may

sP26

301

50/6.0

48/5.2

Present

Present

17Desiccation

-related

protein

Cra

tero

stig

ma

plan

tagi

neum

AAA63

616

33/5.0

34/5.8

Absen

tPresent

Putativeproteiniden

tific

ationan

daccessionnu

mbe

rof

theclosestm

atch

intheNationa

lCen

terforBiotechno

logy

Inform

ation(N

CBI)

databa

seareindicated.App

roximatemolecular

weigh

t(M

W)(kDa)

andisoe

lectricpo

int(p

I)va

lues

asestimated

from

thege

lsareindicated,

alon

gwiththepred

ictedva

lues

oftheclosestmatch

.The

mea

nex

pression

leve

l(�

SD,n

=3)

at65

and

35%

relative

water

conten

t(RW

C)co

mpa

redto

fully

hydrated

plan

tsisindicatedforproteins

that

exhibitedsign

ificant

chan

gesin

abun

danc

e(paired

t-test;P

<0.05

);nsdindicatesno

sign

ificant

differen

cein

expression

.For

proteins

notde

tectab

lein

fully

hydrated

plan

ts(d

eno

voproteins),de

tectionat

65or

35%

RWC

isindicatedby

presen

t(detected)

orab

sent

(not

detected

).



Table 2. Peptide fragments identified by matrix-assisted laser desorption/ionization (MALDI) time-of-flight (TOF)/TOF analysis

Putative identification Peptides matched Score

1 Transketalose ALPTYTPESPADATR (82); KTPSILALSR (28) 1032 F-ATPase (a subunit) LIESPAPGIISR (72); LIESPAPGIISRR (64); IAQIPVSEDYLGR (87);

VINALAKPIDGRGEISASESR (70)292

3 Glu:glyoxylate aminotransferase I DGYPSDPELIFLTDGASK (98); HYLSLTSGGLGAYSDSR (104);IIFTNVGNPHALGQKPLTFPR (38); AKHYLSLTSGGLGAYSDSR (119)

358

4 PSII stability factor HCF136 GFGILDVGYR (42); AAVQETVSATLNR (43) 855 Ascorbate peroxidase TGGPFGTIR (65); QELAHDANNGLDIAVR (54) 1016 psbO GSSFLDPKGR (44); RLTYDEIQSK (67) 1117 psbP EVEYPGQVLR (72); SITDFGSPEEFLSKVDYLLGK (16) 888 Chloroplast FtsH protease VKILQVHSR (20); SYLENQMAVALGGR (36);

AQGGPGGGPGGLGGPMDFGR (58)114

9 GDP-mannose-3′,5′-epimerase QLPIHHIPGPEGVR (70); QLPIHHIPGPEGVRGR (56) 12610 Alcohol dehydrogenase GTFFGNYKPR (60); TLKGTFFGNYKPR (84); IIGVDLNPSRFNEAK (34) 17711 Protein phosphatase type 2C GGFVSNIPGDVPR (39); GGFVSNIPGDVPRVDGQLAVAR (58) 9712 VDAC1.1 SFFTISGEVDTK (84); GPGLYTDIGKK (88) 17013 2-Cys peroxiredoxin SGGLGDLKYPLISDVTK (113); SMKPDPKGSKEYFAAI (65) 17314 dnaK-type molecular chaperone IAGLDVQR (39); EVDEVLLVGGMTR (19); AVITVPAYFNDAQR (31); 8815 RNA-binding protein GSGFVAFSTPEEASR (112); VYVGVFLR (20) 12416 Phosphopyruvate hydratase FRAPVEPY (41); LAKYNQLLR (25); VQIVGDDLLVTNPTR (45) 11117 Desiccation-related protein KLVAGLLAVEAGQDAIIR (35); VEPYGITVAEFTNK (33) 68

The number and amino acid sequences of the peptides matched to proteins in the National Center for Biotechnology Information (NCBI)database are shown. Probability-based MOWSE scores for the individual peptides (ion scores) and an overall score for the protein (derivedfrom the ion scores) derived from Mascot searches are provided. Ion scores > 35 and protein scores > 67 indicate identity or extensivehomology at P < 0.05. M indicates an oxidized Met residue.

25

30

40

55

75

kDa

100% RWC

1

2

3

4

5

7

6

4 5 6 7 pI8

psbP

ADH

100% 65% 35%

Relative water content25

30

40

55

75

kDa

35% RWC

814 15

169

10

17 11

12

13

4 5 6 7 pI8

65% RWC

169

11

(a)

(b)

Figure 2. (a) Changes in proteinabundance in leaf tissue of Xerophytaviscosa during dehydration from 100 to 65and 35% relative water content (RWC).Total protein was isolated, separated byisoelectric focusing (IEF) (non-linear pH3.0–10.0 gradient) followed by sodiumdodecyl sulphate–polyacrylamide gelelectrophoresis (SDS–PAGE) andanalysed using PDQuest software. Threereplicate gels, each with protein isolatedfrom an independent biological replicateconsisting of tissue pooled from threeplants, were run for each RWC.Dehydration-responsive proteins thatwere successfully identified by massspectrometry (MS) analysis arehighlighted on representative gels:increase (square), decrease (circle), denovo (triangle). (b) Enlargement ofrepresentative gels showing examples oftwo late dehydration-responsive proteins,psbP (decreased expression) and alcoholdehydrogenase (ADH) (increasedexpression). The increase in ADHexpression was only statistically significantat 35% RWC.



observed at 35% RWC. GGT1 is localized to the peroxi-some where it catalyses the conversion of glyoxylate toglycine during photorespiration (Liepman & Olsen 2003).As photosynthesis has ceased in X. viscosa at this stage indrying, it is probable that flux through the photorespiratorypathway had also stopped.

In conjunction with the decreased abundance of the fivechloroplastic proteins, there was an increase in levels of achloroplast FtsH protease at both 65 and 35% RWC.Theseproteins constitute a small family of membrane-bound zincmetalloproteases containing an ATPase domain. In Arabi-dopsis, two of the nine chloroplast-targeted FtsH proteaseshave been shown to play a role in the repair of PSII afteroxidative damage.Ftsh2 and to a lesser extent ftsh5 mutantsdisplay elevated photoinhibition because of a reduced turn-over of the D1 polypeptide (Bailey et al. 2002; Sakamotoet al. 2003), and it has been shown that FtsH1 2, 5 and 8form heteromeric complexes, with functional redundancybetween FtsH1 and 5, and between FtsH2 and 8 (Yu, Park& Rodermel 2004; Zaltsman, Ori & Adam 2005). MS/MSanalysis putatively identified the FtsH in the present studyas FtsH1, but the antibody used inWestern analysis (Fig. 3)recognizes all plant FtsH proteins, and so cannot be used todifferentiate between the FtsH isoforms in X. viscosa. Invitro studies have suggested that the FtsH protease complexmay also degrade an unassembled Rieske Fe–S thylakoidprotein (Ostersetzer & Adam 1997), but to date no furthertargets have been identified. An increase in FtsH protease

activity in X. viscosa might serve to repair damage to the D1polypeptide during the early stages of drying when photo-synthesis is still active, but would not be required at lowerRWC. In line with the observation that thylakoid mem-branes are disassembled during drying (Sherwin & Farrant1998; Mundree & Farrant 2000), it is possible that othercomponents of PSII (and other chloroplast proteins) mightbe targeted for degradation by FtsH proteases to halt pho-tosynthesis and prevent ROS formation during waterdeficit. Alternatively, the observation that the ftsh2 andftsh5 mutants are variegated suggests that FtsH proteasesmay play a role in thylakoid formation (Zaltsman et al.2005). Thus, the increase in the FtsH protease observedmight represent synthesis and storage of this protein toallow rapid reassembly of the thylakoid membranes whenwater becomes available again.

Comparison of photosynthetic protein levelsbetween poikilochlorophyllous andhomoiochlorophyllous resurrection plantsduring drying

Xerophyta viscosa is poikilochlorophyllous and dismantlesthylakoid membranes during drying such that only vesicu-lated remnants of thylakoids are present in dry leaves(Sherwin & Farrant 1998; Mundree & Farrant 2000). 2DSDS–PAGE analysis during drying suggested that thesevesicles contain reduced levels of photosynthetic proteins(Fig. 2; Table 1), providing an explanation for the results ofearlier inhibitor studies showing that de novo translation isrequired for resumption of PSII activity during rehydration(Dace et al. 1998). In contrast, Craterostigma spp. arehomoiochlorophyllous, maintaining the integrity of thyla-koid membranes during drying, and Craterostigma wilmsiihas been shown to recover PSII activity upon rehydrationwithout the need for de novo transcription or translation(Cooper & Farrant 2002). This observation suggests thatunlike Xerophyta spp., Craterostigma spp. must maintainand protect PSII proteins during dehydration and in thedesiccated state. To determine whether this is the case,expression levels of three photosynthetic proteins (Lhcb2,psbP and psbS) were monitored during drying in X. viscosaand C. plantagineum. Lhcb2 is a component of the light-harvesting antennae of both photosystem I (PSI) and PSII(Jannson 1999), while psbS is a component of PSII thatplays a role in non-photochemical quenching (Niyogi et al.2005). In X. viscosa, levels of all three proteins werereduced at 55% RWC, and by 30% RWC psbP and Lhcb2were no longer detectable (Fig. 4). Thus, de novo synthesisof these two proteins at least would likely be requiredduring rehydration for resumption of PSII activity. In con-trast, psbP and psbS expression remained constant through-out dehydration in C. plantagineum, while Lhcb2 levelsdecreased but were still detectable at 11% RWC (Fig. 4).Thus, poikilochlorophylly in X. viscosa involves the break-down of photosynthetic proteins during dismantling of thethylakoid membranes, while in the homoiochlorophyllousC. plantagineum PSII protein levels appear to be largely

FtsH (70 kDa)

2-Cys-Prx (25 kDa)

HCF136 (37 kDa)

psbP (25 kDa)

100 65 35 RWC (%)

Ponceau stain

Up

Down

Figure 3. Validation of protein expression data by Western blotanalysis. Antibodies were obtained for two proteins showingincreased expression during dehydration according to PDQuestanalysis (FtsH protease and 2-Cys-Prx) and two showing reducedexpression (HCF136 and psbP). Total protein (30 mg) wasseparated by one-dimensional (1D) sodium dodecylsulphate–polyacrylamide gel electrophoresis (SDS–PAGE),transferred to nitrocellulose membrane and probed with theappropriate antibody. Equal loading of the gel was verified byPonceau staining of the membrane after protein transfer, andproteins were sized by comparison to the migration of bands in aprestained protein ladder. The results shown are from oneexperiment representative of three carried out on independentbiological replicates.



maintained during drying. In light of this difference, analysisof FtsH protein levels was also carried out. In X. viscosa,FtsH protein levels again increased early on during drying(at 70% RWC) and remained elevated throughout, while inC. plantagineum FtsH levels remained unchanged until 11%RWC (Fig. 4). Whether increased FtsH levels in X. viscosaare causally related to the increased degradation of chloro-plast proteins during drying in is unknown. Identification ofthe proteins targeted for FtsH-mediated degradation in X.viscosa would assist in determining whether this proteasecomplex plays a role in the controlled shutdown of photo-synthesis that occurs during drying in this species.

Proteins involved in protection againstoxidative stress

Three proteins with a potential role in protection againstoxidative damage changed in abundance in X. viscosaduring drying: 2-Cys-Prx and GDP-mannose-3′, 5′-epimerase increased and ascorbate peroxidase (APX)declinedduring drying (Table 1).2-Cys-Prxs are chloroplast-targeted proteins that reduce peroxide substrates to thecorresponding alcohol. It is thought that they function toprotect chloroplasts against ROS generated during photo-synthesis, as antisense Arabidopsis plants displayedincreased levels of photoinhibition (Baier & Dietz 1999).

2-Cys-Prx protein levels increase moderately in high light inArabidopsis, in response to a redox signal correlating withthe availability of electron acceptors downstream of PSI(Horling et al. 2003; Baier, Stroher & Dietz 2004). Similarly,increased electron pressure in the electron transport chaincaused by inhibition of photosynthesis during water stressmight result in an increase in 2-Cys-Prx expression in X.viscosa, andmay help to protect against oxidative damage tothe chloroplast.

Ascorbate acts as a scavenger of H2O2 and is converted tomonodehydroascorbate in a reaction catalysed by APX.Increased activities of the antioxidant enzymes APX, glu-tathione reductase and superoxide dismutase have beenreported in several resurrection plants including X. viscosaduring drying (Sherwin & Farrant 1998; Farrant 2000). Inthe present study, a ninefold decrease in the abundance of acytosolic isoform of APX was detected at 35% RWC. APXconstitutes a multigene family, with eight members in Ara-bidopsis (Panchuk, Zentgraf & Volkov 2005), and it is pos-sible that differential regulation of the various isoformsmay occur during drying in X. viscosa. However, it has beensuggested that while APX plays a role in abiotic stress tol-erance in vegetative tissues, it may not be involved in theacquisition of desiccation tolerance in seeds (Bailly 2004).Because the mechanism of desiccation tolerance in Xero-phyta species has been proposed to involve the induction of‘seed-specific’ genes in vegetative tissues during the latterstages of drying (Illing et al. 2005), it is possible that APX isnot involved in ROS scavenging at low water contents inleaf tissue of X. viscosa.An indication of a possible increase in ascorbate synthe-

sis in X. viscosa during the early stages of drying was theapproximately twofold increased abundance of GDP-mannose-3′,5′-epimerase observed at 65 (and 35) % RWC.This enzyme catalyses the conversion of GDP-mannose toGDP-l-galactose or GDP-l-gulose, and represents the firststep in the de novo synthesis of ascorbate (Wolucka et al.2001). It has recently been demonstrated that applicationof methyl jasmonate (a signalling molecule involved inresponses to biotic and abiotic stress) leads to transcrip-tional up-regulation of the gene encoding GDP-mannose-3′,5′-epimerase and increased ascorbate biosynthesis intobacco cell cultures (Wolucka, Goossens & Inze 2005).Correspondingly, measurement of total ascorbate contentrevealed a 25% increase in ascorbate levels in X. viscosaplants at 65% RWC compared to 100% RWC (Fig. 5).However, this increase in the ascorbate pool was transient,because at 35% RWC ascorbate content was not signifi-cantly different to that observed at 100% RWC, althoughGDP-mannose-3′,5′-epimerase protein levels did notdecline. This may indicate that increased ascorbate pro-duction is an ‘early’ response similar to that previouslyreported in vegetative tissues of non-resurrection plants inresponse to a variety of biotic and abiotic stresses (Woluckaet al. 2005). In Arabidopsis GDP-mannose-3′,5′-epimerasecopurifies with a HSP70 protein, which is believed to beinvolved in folding or regulation of the enzyme (Wolucka &Van Montagu 2003). Interestingly, a dnaK-type HSP70 was

Lhcb2

psbS

psbP

FtsH

100

70 45 11

10070

55 30

X. viscosa C. plantagineum

% RWC

Ponceau stain

Figure 4. Differential expression of photosynthetic proteinsbetween Xerophyta viscosa and Craterostigma plantagineumduring dehydration. Plants were dried by withholding water, andsamples were collected at various relative water content (RWC).Total protein (30 mg) was separated by one-dimensional (1D)sodium dodecyl sulphate–polyacrylamide gel electrophoresis(SDS–PAGE), transferred to nitrocellulose membrane andprobed with antibodies to psbP, psbS and Lhcb2. Equal loadingof the gel was verified by Ponceau staining of the membraneafter protein transfer. The results shown are from oneexperiment representative of three carried out on independentbiological replicates.



also detected in X. viscosa at 65 and 35% RWC, but not at100% RWC.

Levels of two glycolytic enzymes increaseduring drying

Two glycolytic enzymes increased in abundance duringdrying of X. viscosa: phosphopyruvate hydratase andalcohol dehydrogenase (ADH). Carbohydrate metabolismis modulated in X. viscosa and other resurrection plantsduring drying, particularly towards the synthesis of sucrose(Bianchi et al. 1991; Whittaker et al. 2001), and possiblytowards the synthesis of compatible solutes such as sorbitol(Mundree et al. 2000). The accumulation of sucrose isindependent of photosynthetic activity because it occurspredominately at lower RWC after the cessation of photo-synthesis (Cooper & Farrant 2002).The function of sucroseaccumulation remains unclear, but it has been suggestedthat it may act as a water replacement molecule to stabilizemembrane and protein structure or in vitrification ofthe cytoplasm in desiccated tissue, and as an energysource during rehydration (Hoekstra et al. 2001; Buitink &Leprince 2004; Vicré et al. 2004a). Phosphopyruvatehydratase (enolase) catalyses the conversion of 2-phospho-d-glycerate to phosphoenolpyruvate during glycolysis, andcatalyses the reverse reaction in gluconeogenesis. Anincrease in flux through the gluconeogenic pathway duringdrying would provide an increased pool of hexose phos-phate substrates required for both sucrose and sorbitol syn-thesis. ADH catalyses the conversion of ethanal to ethanolin anaerobic glycolysis, and the ADH gene in Arabidopsis isup-regulated in abscisic acid (ABA)-dependent mannerin response to dehydration (de Bruxelles et al. 1996).

However, the function of this enzyme is unclear exceptunder anoxic conditions such as flooding.

Detection of an RNA-binding protein at35% RWC

An RNA-binding protein was detected in soluble proteinextracts from X. viscosa at 35%RWC but not at 65 or 100%RWC. The subcellular localization of the closest match(AAK30205) is not known, but a putative chloroplasttransit peptide is predicted by ChloroP analysis (Emanuels-son, Nielsen & von Heijne 1999). Messenger RNAsrequired for the repair of dehydration-induced damage arestored in desiccated T. ruralis plants (Dhindsa & Bewley1978), and there is evidence from inhibitor studies tosuggest that mRNAs may also be stored in Xerophytaspecies during desiccation (Dace et al. 1998). In addition,the maintenance of two mRNAs encoding components ofPSII: psbA (D1 polypeptide) and psbP has been demon-strated in desiccated X. humilis leaf tissue (Collett et al.2003). Microarray analysis has previously identified a chlo-roplast ribonucleoprotein (cpRNP) homologous to cp29Afrom tobacco as up-regulated in X. humilis at RWCs below50% (Collett et al. 2004). CpRNPs have been shown to bindand stabilize psbA mRNA in tobacco stromal extracts(Nakamura et al. 2001), and it is possible that RNA-bindingproteins may serve to stabilize specific mRNA transcripts,for example,mRNAs for PSII proteins during desiccation inXerophyta species. RNA co-immunoprecipitation assays onsamples collected from Xerophyta plants at different stagesof drying would allow the complement of RNAs bound bythese proteins at different RWC to be examined.

Increased expression of a phosphataseimplicated in ABA-signalling

The expression of a type 2C phosphatase increased approxi-mately fourfold at both 65 and 35% RWC. These proteinsact as negative regulators of signalling pathways by oppos-ing the action of protein kinases, and have been implicatedas negative regulators of theABA signalling.ABA is one ofthe major signalling molecules involved in response tochanges in water availability in plants (Himmelbach, Yang& Grill 2003). ABA concentrations increase in response toosmotic stress, leading to the modulation of gene expressionthroughABA-response elements found in the promoters ofmany dehydration-induced genes in a wide range of speciesincluding resurrection plants, as well as regulating transla-tion and protein turnover (Himmelbach et al. 2003). In Ara-bidopsis, the type 2C phosphatasesABI1 and 2 are negativeregulators of ABA signalling and are up-regulated byABAat the mRNA level (Gosti et al. 1999). It is thought that sucha negative feedback loop allows the plant to reset theABAsignalling pathway in order to continually monitor thepresence or absence of ABA (Gosti et al. 1999).

CONCLUSIONS

In summary, analysis of the proteome of X. viscosa identi-fied a number of dehydration-responsive proteins that may

0

4

8

12

16

20

100 65 35

Relative water content (%)

*

mmol

Asc

orba

teg

–1dr

y bi

omas

s

Figure 5. The ascorbate pool of Xerophyta viscosa increasestransiently during dehydration. Total ascorbate content wasmeasured after reduction of dehydroascorbate to ascorbate bydithiothreitol (DTT). Mean ascorbate content from threeindependent biological replicates (mmol g-1 dry biomass) � SD isindicated. The * indicates an ascorbate content significantlydifferent (t-test, P < 0.05) to that of fully hydrated plants.



play a role in desiccation tolerance in this species, includingantioxidants, an RNA-binding protein and a number ofproteins involved in photosynthesis. These proteins couldbe grouped into three classes based on their expres-sion pattern: early-dehydration, late-dehydration and full-dehydration responsive. Proteins that are late-dehydrationresponsive may be of greatest interest in uncovering themolecular basis of desiccation tolerance, because suchchanges in expression are likely unique to resurrectionplants; non-desiccation tolerant species such as Arabidopsiscannot survive at such low RWC. This study also demon-strated that protein sequence information in the NCBIdatabase could be used to identify proteins from plantspecies that are phylogenetically distant from model plants,albeit with a relatively modest success rate (approximately30%). Identification of these proteins provides a foundationfor further functional studies to determine their precisebiochemical roles in desiccation tolerance. Proteomicanalysis has several advantages over transcriptomics in thatit provides a view of the end-point of gene expression, i.e.the actively translated component of the mRNA pool, andallows the investigation of post-transcriptional and post-translational regulatory events.However, it should be notedthat the 2D SDS–PAGE approach used here has limita-tions, as it allows the detection and analysis of only a subsetof relatively abundant and soluble proteins. One possiblemethod to increase the number of proteins that can bestudied by 2D SDS–PAGE is the fractionation of the pro-teome into subcellular fractions. Analysis of the X. viscosachloroplast proteome during dehydration would be of par-ticular interest given that chloroplast proteins constituted 8of the 17 dehydration-responsive proteins identified bytandem MS, and that reduction of PSII proteins duringdrying seems to be a component of poikilochlorophylly.Alternatively, the gel-free iTRAQ MS system could beemployed to greatly increase the number of proteins thatcan be analysed, with the additional benefit that hydropho-bic proteins lost during isoelectric focusing (IEF) can bestudied (Suzuki, Simon & Slabas 2006). A systems biologyapproach combining proteomics, metabolomics and tran-scriptomics offers the possibility to better understand theability of resurrection plants to tolerate desiccation, and todetermine at which level of control these changes areeffected.

ACKNOWLEDGMENTS

We are grateful to the following people for the donation ofantibodies: psbP, Colin Robinson (University of Warwick);HCF136, Peter Westhoff (Heinrich-Heine University, Dus-seldorf); FtsH, Zach Adams (Hebrew University of Jerusa-lem); 2-Cys Prx, Karl Dietz (University of Bielefeld); andLchb2/psbS, Helen Collett (University of Cape Town).Thanks also to Lee Sweetlove (University of Oxford) foraccess to 2D SDS–PAGE equipment, Peter Gehrig (Func-tional Genomics Centre, Zurich) for excellent technicalsupport with MS analysis, Katherine Denby (University ofWarwick) for useful discussions on statistical analysis, and

Enrico Martinoia (University of Zurich). R.A.I. wassupported by a fellowship from the Claude Leon Founda-tion, South Africa; U.G.S. was supported by the projectNovel Ion Channels in Plants (EU HPRN-CT-00245); andS.G.M. was funded by the National Research Foundation,South Africa.

REFERENCES

Alpert P. (2006) Constraints of tolerance: why are desiccation-tolerant organisms so small or rare? Journal of ExperimentalBotany 209, 1575–1584.

Baier M. & Dietz K.-J. (1999) Protective function of chloroplast2-cysteine peroxiredoxin in photosynthesis. Evidence fromtransgenic Arabidopsis. Plant Physiology 119, 1407–1414.

Baier M., Stroher E. & Dietz K.-J. (2004) The acceptor availabilityat photosystem I and ABA control nuclear expression of 2-Cysperoxiredoxin-A in Arabidopsis thaliana. Plant & Cell Physiol-ogy 45, 997–1006.

Bailey S., Thompson E., Nixon P.J., Horton P., Mullineaux C.W.,Robinson C. & Mann N.H. (2002) A critical role for the Var2FtsH homologue of Arabidopsis thaliana in the photosystem IIrepair cycle in vivo. Journal of Biological Chemistry 277, 2006–2011.

Bailly C. (2004) Active oxygen species and antioxidants in seedbiology. Seed Science Research 14, 93–107.

Bartels D., Schneider K., Terstappen G., Piatkowski D. & SalaminiF. (1990) Molecular cloning of abscisic acid-modulated geneswhich are induced during desiccation of the resurrection plantCraterostigma plantagineum. Planta 181, 27–34.

Bernacchia G. & Furini A. (2004) Biochemical and molecularresponses to water stress in resurrection plants. PhysiologiaPlantarum 121, 175–181.

Bernacchia G., Salamini F. & Bartels D. (1996) Molecular charac-terization of the rehydration process in the resurrection plantCraterostigma plantagineum. Plant Physiology 111, 1043–1050.

Bianchi G., Gamba A., Murelli C., Salamini F. & Bartels D. (1991)Novel carbohydrate metabolism in the resurrection plantCraterostigma plantagineum. The Plant Journal 1, 355–359.

Blomstedt C.K.,Gianello R.D.,Gaff D.F.,Hamill J.D. &Neale A.D.(1998) Differential gene expression in desiccation-tolerant anddesiccation-sensitive tissue of the resurrection grass, Sporobolusstapfianus. Australian Journal of Plant Physiology 25, 937–946.

Bockel C., Salamini F. & Bartels D. (1998) Isolation and character-ization of genes expressed during early events of the dehydrationprocess in the resurrection plant Craterostigma plantagineum.Journal of Plant Physiology 152, 158–166.

de Bruxelles G.L., Peacock W.J., Dennis E.S. & Dolferus R. (1996)Abscisic acid induces the alcohol dehydrogenase gene in Arabi-dopsis. Plant Physiology 111, 381–391.

Buitink J. & Leprince O. (2004) Glass formation in plant anhydro-biotes: survival in the dry state. Cryobiology 48, 215–228.

Collett H., Butowt R., Smith J., Farrant J. & Illing N. (2003) Pho-tosynthetic genes are differentially transcribed during thedehydration–rehydration cycle in the resurrection plant, Xero-phyta humilis. Journal of Experimental Botany 54, 2593–2595.

Collett H., Shen A., Gardner M., Farrant J.M., Denby K.J. & IllingN. (2004) Towards transcript profiling of desiccation tolerance inXerophyta humilis: construction of a normalized 11 k X. humiliscDNA set and microarray expression analysis of 424 cDNAs inresponse to dehydration. Physiologia Plantarum 122, 39–53.

Cooper K. & Farrant J.M. (2002) Recovery of the resurrectionplant Craterostigma wilmsii from desiccation: protection versusrepair. Journal of Experimental Botany 53, 1805–1813.

Dace H., Sherwin H.W., Illing N. & Farrant J.M. (1998) Use of



metabolic inhibitors to elucidate mechanisms of recovery fromdesiccation stress in the resurrection plant Xerophyta humilis.Plant Growth Regulation 24, 171–177.

Dhindsa R. & Bewley J. (1978) Messenger RNA is conservedduring drying of the drought-tolerant moss Tortula ruralis. Pro-ceedings of the National Academy of Science USA 75, 842–846.

Emanuelsson O., Nielsen H. & von Heijne G. (1999) ChloroP, aneural network-based method for predicting chloroplast transitpeptides and their cleavage sites. Protein Science 8, 978–984.

Farrant J.M. (2000) A comparison of mechanisms of desiccationtolerance among three angiosperm resurrection plant species.Plant Ecology 151, 29–39.

Gaff D. (1971) Desiccation-tolerant plants in Southern Africa.Science 174, 1033–1034.

Gosti F., Beaudoin N., Serizet C., Webb A.A.R., Vartanian N. &Giraudat J. (1999) ABI1 protein phosphatase 2C is a negativeregulator of abscisic acid signaling. Plant Cell 11, 1897–1910.

Gygi S.P., Rochon Y., Franza B.R. & Aebersold R. (1999) Correla-tion between protein and mRNA abundance in yeast. Molecularand Cellular Biology 19, 1720–1730.

Hakala M., Tuominen I., Keranen M., Tyystjarvi T. & Tyystjarvi E.(2005) Evidence for the role of the oxygen-evolving manganesecomplex in photoinhibition of photosystem II. Biochimica etBiophysica Acta (BBA) – Bioenergetics 1706, 68–80.

Himmelbach A., Yang Y. & Grill E. (2003) Relay and control ofabscisic acid signaling. Current Opinion in Plant Biology 6, 470–479.

Hoekstra F.A., Golovina E.A., Tetteroo F.A.A. & Wolkers W.F.(2001) Induction of desiccation tolerance in plant somaticembryos: how exclusive is the protective role of sugars? Cryobi-ology 43, 140–150.

Horling F., Lamkemeyer P., Konig J., Finkemeier I., KandlbinderA., Baier M. & Dietz K.-J. (2003) Divergent light-, ascorbate-,and oxidative stress-dependent regulation of expression of theperoxiredoxin gene family in Arabidopsis. Plant Physiology 131,317–325.

Ideker T., Thorsson V., Ranish J.A., Christmas R., Buhler J.,Eng J.K., Bumgarner R., Goodlett D.R., Aebersold R. &Hood L. (2001) Integrated genomic and proteomic analyses of asystematically perturbed metabolic network. Science 292, 929–934.

Illing N.,Denby K.J., Collett H., Shen A. & Farrant J.M. (2005) Thesignature of seeds in resurrection plants: a molecular and physi-ological comparison of desiccation tolerance in seeds and veg-etative tissues. Integrative & Comparative Biology 45, 771–787.

Ingle R., Smith J. & Sweetlove L. (2005) Responses to nickel in theproteome of the hyperaccumulator plant Alyssum lesbiacum.Biometals 18, 627–641.

Ingram J. & Bartels D. (1996) The molecular basis of dehydrationtolerance in plants.Annual Review of Plant Physiology and PlantMolecular Biology 47, 377–403.

Ingram J., Chandler J.W., Gallagher L., Salamini F. & Bartels D.(1997) Analysis of cDNA clones encoding sucrose-phosphatesynthase in relation to sugar interconversions associated withdehydration in the resurrection plant Craterostigma plan-tagineum Hochst. Plant Physiology 115, 113–121.

Jannson S. (1999) A guide to the Lhc genes and their relatives inArabidopsis. Trends in Plant Science 4, 236–240.

Liepman A.H. & Olsen L.J. (2003) Alanine aminotransferasehomologs catalyze the glutamate:glyoxylate aminotransferasereaction in peroxisomes of Arabidopsis. Plant Physiology 131,215–227.

Meurer J., Plücken H.,Kowallik K. &Westhoff P. (1998) A nuclear-encoded protein of prokaryotic origin is essential for the stabilityof photosystem II in Arabidopsis thaliana. European MolecularBiology Organisation Journal 17, 5286–5297.

Millar A.H., Sweetlove L.J., Giege P. & Leaver C.J. (2001) Analysisof the Arabidopsis mitochondrial proteome. Plant Physiology127, 1711–1727.

Mittler R., Merquiol E., Hallak-Herr E., Rachmilevitch S., KaplanA. & Cohen M. (2001) Living under a ‘dormant’ canopy: amolecular acclimation mechanism of the desert plant Retamaraetam. The Plant Journal 25, 407–416.

Moore J., Westall K., Ravenscroft N., Farrant J., Lindsey G. &Brandt W. (2005) The predominant polyphenol in the leaves ofthe resurrection plant Myrothamnus flabellifolius, 3,4,5 tri-O-galloylquinic acid, protects membranes against desiccation andfree radical-induced oxidation. Biochemical Journal 385, 301–308.

Moore J.P., Nguema-Ona E., Chevalier L., Lindsey G.G., BrandtW.F., Lerouge P., Farrant J.M. & Driouich A. (2006) Response ofthe leaf cell wall to desiccation in the resurrection plant Myroth-amnus flabellifolius. Plant Physiology 141, 651–662.

Mowla S.B., Thomson J.A., Farrant J.M. & Mundree S.G. (2002) Anovel stress-inducible antioxidant enzyme identified from theresurrection plant Xerophyta viscosa Baker.Planta 215, 716–726.

Müller J., Boller T. & Wiemken A. (1995) Trehalose and trehalasein plants: recent developments. Plant Science 112, 1–9.

Mundree S.G. & Farrant J.M. (2000) Some physiological andmolecular insights into the mechanisms of desiccation tolerancein the resurrection plant Xerophyta viscosa Baker. In PlantTolerance to Abiotic Stresses in Agriculture: Role of GeneticEngineering (ed. J. Cherry), pp. 201–222. KluwerAcademic Pub-lishers, Dordrecht, the Netherlands.

Mundree S.G., Whittaker A., Thomson J.A. & Farrant J.M. (2000)An aldose reductase homolog from the resurrection plant Xero-phyta viscosa Baker. Planta 211, 693–700.

Nakamura T., Ohta M., Sugiura M. & Sugita M. (2001) Chloroplastribonucleoproteins function as a stabilizing factor of ribosome-free mRNAs in the stroma. Journal of Biological Chemistry 276,147–152.

Neale A.D., Blomstedt C.K., Bronson P., Le T.N., Guthridge K.,Evans J., Gaff D.F. & Hamill J.D. (2000) The isolation of genesfrom the resurrection grass Sporobolus stapfianus which areinduced during severe drought stress.Plant, Cell & Environment23, 265–277.

Niyogi K.K., Li X.-P., Rosenberg V. & Jung H.-S. (2005) Is PsbS thesite of non-photochemical quenching in photosynthesis? Journalof Experimental Botany 56, 375–382.

Ohnishi N., Allakhverdiev S.I., Takahashi S., Higashi S., WatanabeM., Nishiyama Y. & Murata N. (2005) Two-step mechanism ofphotodamage to photosystem II: step 1 occurs at the oxygen-evolving complex and step 2 occurs at the photochemical reac-tion center. Biochemistry 44, 8494–8499.

Oliver M.J. & Bewley J. (1984) Plant desiccation and protein syn-thesis:VI. Changes in protein synthesis elicited by desiccation ofthe moss Tortula ruralis are effected at the translational level.Plant Physiology 74, 923–927.

Oliver M.J. & Bewley J. (1997) Desiccation tolerance of planttissues: a mechanistic overview. Horticultural Reviews 18, 171–214.

Oliver M.J., Tuba Z. & Mishler B. (2000) The evolution of vegeta-tive desiccation tolerance in land plants. Plant Ecology 151,85–100.

Oliver M.J., Dowd S.E., Zaragoza J., Mauget S.A. & Payton P.R.(2004) The rehydration transcriptome of the desiccation-tolerant bryophyte Tortula ruralis: transcript classification andanalysis. BMC Genomics 5, 89.

Ostersetzer O. & Adam Z. (1997) Light-stimulated degradation ofan unassembled Rieske FeS protein by a thylakoid-bound pro-tease: the possible role of the FtsH protease. Plant Cell 9, 957–965.



Panchuk I., Zentgraf U. & Volkov R. (2005) Expression of the Apxgene family during leaf senescence of Arabidopsis thaliana.Planta 222, 926–932.

Röhrig H., Schmidt J., Colby T., Brautigam A., Hufnagel P. &Bartels D. (2006) Desiccation of the resurrection plant Cra-terostigma plantagineum induces dynamic changes in proteinphosphorylation. Plant, Cell & Environment 29, 1606–1617.

Sakamoto W., Zaltsman A., Adam Z. & Takahashi Y. (2003) Coor-dinated regulation and complex formation of YELLOW VAR-IEGATED1 and YELLOW VARIEGATED2, chloroplasticFtsH metalloproteases involved in the repair cycle of photosys-tem II in Arabidopsis thylakoid membranes.Plant Cell 15, 2843–2855.

Sherwin H.W. & Farrant J.M. (1996) Differences in rehydration ofthree desiccation-tolerant angiosperm species.Annals of Botany78, 703–710.

Sherwin H.W. & Farrant J.M. (1998) Protection mechanismsagainst excess light in the resurrection plants Craterostigmawilmsii and Xerophyta viscosa.Plant Growth Regulation 24, 203–210.

Smirnoff N. (1993) The role of active oxygen in the response ofplants to water deficit and desiccation. New Phytologist 125,27–58.

Suzuki I., Simon W.J. & Slabas A.R. (2006) The heat shockresponse of Synechocystis sp. PCC 6803 analysed by transcrip-tomics and proteomics. Journal of Experimental Botany 57,1573–1578.

VanderWilligen C., Pammenter N.W.,Mundree S.G. & Farrant J.M.(2004) Mechanical stabilization of desiccated vegetative tissuesof the resurrection grass Eragrostis nindensis: does a TIP 3;1and/or compartmentalization of subcellular components andmetabolites play a role? Journal of Experimental Botany 55,651–661.

Vicré M., Farrant J.M. & Driouich A. (2004a) Insights into thecellular mechanisms of desiccation tolerance among angiospermresurrection plant species. Plant, Cell & Environment 27, 1329–1340.

Vicré M., Lerouxel O., Farrant J.M., Lerouge P. & Driouich A.(2004b) Composition and desiccation-induced alterations of the

cell wall in the resurrection plant Craterostigma wilmsii. Physi-ologia Plantarum 120, 229–239.

Wang S.Y., Jiao H.J. & Faust M. (1991) Changes in ascorbate,glutathione, and related enzyme activities during thidiazuron-induced bud break of apple. Physiologia Plantarum 82, 231–236.

Whittaker A., Bochicchio A., Vazzana C., Lindsey G. & Farrant J.(2001) Changes in leaf hexokinase activity and metabolite levelsin response to drying in the desiccation-tolerant speciesSporobolus stapfianus and Xerophyta viscosa. Journal of Experi-mental Botany 52, 961–969.

Wolucka B.A. & Van Montagu M. (2003) GDP-mannose 3′,5′-epimerase forms GDP-l-gulose, a putative intermediate for thede novo biosynthesis of vitamin C in plants. Journal of BiologicalChemistry 278, 47483–47490.

Wolucka B.A., Persiau G., Van Doorsselaere J., Davey M.W.,Demol H., Vandekerckhove J., Van Montagu M., Zabeau M. &Boerjan W. (2001) Partial purification and identification ofGDP-mannose 3′,5′-epimerase of Arabidopsis thaliana, a keyenzyme of the plant vitamin C pathway. Proceedings of theNational Academy of Science USA 98, 14843–14848.

Wolucka B.A., Goossens A. & Inze D. (2005) Methyl jasmonatestimulates the de novo biosynthesis of vitamin C in plant cellsuspensions. Journal of Experimental Botany 56, 2527–2538.

Wood A.J., Duff R.J. & Oliver M.J. (1999) Expressed sequencetags (ESTs) from desiccated Tortula ruralis identify a largenumber of novel plant genes. Plant & Cell Physiology 40, 361–368.

Yu F., Park S. & Rodermel S.R. (2004) The Arabidopsis FtsHmetalloprotease gene family: interchangeability of subunits inchloroplast oligomeric complexes. The Plant Journal 37, 864–876.

Zaltsman A.,Ori N. & Adam Z. (2005) Two types of FtsH proteasesubunits are required for chloroplast biogenesis and photosys-tem II repair in Arabidopsis. Plant Cell 17, 2782–2790.

Received 25 October 2006; accepted for publication 14 November2006



proteomic analysis of leaf proteins during dehydration of the resurrection plant xerophyta viscosa

Documents